The present invention relates to nucleic acid constructs, pharmaceutical compositions and methods which can be used to regulate angiogenesis in specific tissue regions of a subject. More particularly, the present invention relates to isolated polynucleotide sequences exhibiting endothelial cell specific promoter activity, and methods of use thereof and, yet more particularly, to a modified-preproendothelin-1 (PPE-1) promoter which exhibits increased activity and specificity in endothelial cells, and nucleic acid constructs, which can be used to either activate cytotoxicity in specific cell subsets, thus, enabling treatment of diseases characterized by aberrant neovasculariza tion or cell growth or induce the growth of new blood vessels, thus, enabling treatment of ischemic diseases. The invention further relates to modifications of the PPE promoter, which enhance its expression in response to physiological conditions including hypoxia and angiogenesis, and novel angiogenic endothelial-specific combined therapies.
Angiogenesis:
Angiogenesis is the growth of new blood vessels, a process that depends mainly on locomotion, proliferation, and tube formation by capillary endothelial cells. During angiogenesis, endothelial cells emerge from their quiescent state and proliferate rapidly. Although the molecular mechanisms responsible for transition of a cell to angiogenic phenotype are not known, the sequence of events leading to the formation of new vessels has been well documented [Hanahan, D., Science 277, 48-50, (1997)]. The vascular growth entails either endothelial sprouting [Risau, W., Nature 386, 671-674, (1997)] or intussusceptions [Patan, S., et al; Microvasc. Res. 51, 260-272, (1996)]. In the first pathway, the following sequence of events may occur: (a) dissolution of the basement of the vessel, usually a post capillary venule, and the interstitial matrix; (b) migration of endothelial cells toward the stimulus; (c) proliferation of endothelial cells trailing behind the leading endothelial cell (s); (d) formation of lumen (canalization) in the endothelial array/sprout; (e) formation of branches and loops by confluencial anastomoses of sprouts to permit blood flow; (f) investment of the vessel with pericytes (i.e., periendothelial cells and smooth muscle cells); and (g) formation of basement membrane around the immature vessel. New vessels can also be formed via the second pathway: insertion of interstitial tissue columns into the lumen of preexisting vessels. The subsequent growth of these columns and their stabilization result in partitioning of the vessel lumen and remodeling of the local vascular network.
Angiogenesis occurs under conditions of low oxygen concentration (ischemia and tumor metastases etc.) and thus may be an important environmental factor in neovascularization. The expression of several genes including erythropoietin, transferrin and its receptor, most of glucose transport and glycolytic pathway genes, LDH, PDGF-BB, endothelin-1 (ET-1), VEGF and VEGF receptors is induced under hypoxic conditions by the specific binding of the Hypoxia Inducible Factor (HIF-1) to the Hypoxic Response Element (HRE) regulating the transcription of these genes. Expression of these genes in response to hypoxic conditions enables the cell to function under low oxygen conditions.
The angiogenic process is regulated by angiogenic growth factors secreted by tumor or normal cells as well as the composition of the extracellular matrix and by the activity of endothelial enzymes (Nicosia and Ottinetti, 1990, Lab. Invest., 63, 115). During the initial stages of angiogenesis, endothelial cell sprouts appear through gaps in the basement membrane of pre-existing blood vessels (Nicosia and Ottinetti, 1990, supra; Schoefl, 1963, Virehous Arch, Pathol. Anat. 337, 97-141; Ausprunk and Folkman, 1977, Microvasc. Res. 14, 53-65; Paku and Paweletz, 1991, Lab. Invest. 63, 334-346). As new vessels form, their basement membrane undergoes complex structural and compositional changes that are believed to affect the angiogenic response (Nicosia, et. al., 1994, Exp Biology. 164, 197-206).
Angiogenesis and Pathology:
A variety of angiogenic factors govern the angiogenic process. It is understood that during pathology, the fine balance between pro-angiogenic factors and anti-angiogenic factors is disrupted, thereby eliciting nonself-limiting endothelial and periendothelial cell-proliferation. Although angiogenesis is a highly regulated process under normal conditions, many diseases (characterized as “angiogenic diseases”) are driven by persistent unregulated angiogenesis. In such disease states, unregulated angiogenesis can either cause a particular disease directly or exacerbate an existing pathological condition. For example, ocular neovascularization has been implicated as the most common cause of blindness and underlies the pathology of approximately twenty diseases of the eye. In certain previously existing conditions such as arthritis, newly formed capillary blood vessels invade the joints and destroy cartilage. In diabetes, new capillaries formed in the retina invade the vitreous humor, causing bleeding and blindness. Until recently, the angiogenesis that occurs in diseases of ocular neovascularization, arthritis, skin diseases, and tumors, had been difficult to suppress therapeutically.
Unbalanced angiogenesis typifies various pathological conditions and often sustains progression of the pathological state. For example, in solid tumors, vascular endothelial cells divide about 35 times more rapidly than those in normal tissues (Denekamp and Hobson, 1982 Br. J. Cancer 46:711-20). Such abnormal proliferation is necessary for tumor growth and metastasis (Folkinan, 1986 Cancer Res. 46:467-73).
Vascular endothelial cell proliferation is also important in chronic inflammatory diseases such as rheumatoid arthritis, psoriasis and synovitis, where these cells proliferate in response to growth factors released within the inflammatory site (Brown & Weiss, 1988, Ann. Rheum. Dis. 47:881-5).
In atherosclerosis, formation of an atherosclerotic plaque is triggered by a monoclonal expansion of endothelial cells in blood vessels (Alpem-Elran 1989, J. Neurosurg. 70:942-5). Furthermore, in diabetic retinopathy, blindness is thought to be caused by basement membrane changes in the eye, which stimulate uncontrolled angiogenesis and consumption of the retina (West and Kumar, 1988, Lancet 1:715-6).
Endothelial cells are also involved in graft rejection. In allograft rejection episodes, endothelial cells express pro-adhesive determinants that direct leukocyte traffic to the site of the graft. It is believed that the induction of leukocyte adhesion molecules on the endothelial cells in the graft may be induced by locally-released cytokines, as is known to occur in an inflammatory lesion.
Abrogated angiogenesis, on the other hand, is also a major factor in disease development, such as in atherosclerosis induced coronary artery blockage (e.g., angina pectoris), in necrotic damage following accidental injury or surgery, or in gastrointestinal lesions such as ulcers.
Hence, regulating or modifying the angiogenic process can have an important therapeutic role in limiting the contributions of this process to pathological progression of an underlying disease state as well as providing a valuable means of studying their etiology.
Recently significant progress in the development of endothelial regulating agents, whether designed to be inhibitory or stimulatory, has been made. For example, administration of βFGF protein, within a collagen-coated matrix, placed in the peritoneal cavity of adult rats, resulted in a well-vascularized and normally perfused structure (Thompson, et al., PNAS 86:7928-7932, 1989). Injection of βFGF protein into adult canine coronary arteries during coronary occlusion reportedly led to decreased myocardial dysfunction, smaller myocardial infarctions, and increased vascularity (Yanagisawa-Miwa, et al., Science 257:1401-1403, 1992). Similar results have been reported in animal models of myocardial ischemia using βFGF protein (Harada, et al., J Clin Invest 94:623-630, 1994, Unger, et al., Am J Physiol 266:H1588-H1595, 1994).
However, for mass formation of long lasting functional blood vessel there is a need for repeated or long term delivery of the above described protein factors, thus limiting their use in clinical settings. Furthermore, in addition to the high costs associated with the production of angiogenesis-regulating factors, efficient delivery of these factors requires the use of catheters to be placed in the coronary arteries, which further increases the expense and difficulty of treatment.
Therefore, the fundamental goal of all anti-angiogenic therapy is to return foci of proliferating microvessels to their normal resting state, and to prevent their regrowth [Cancer: Principles & Practice of Oncology, Fifth Edition, edited by Vincent T. DeVita, Jr., Samuel Hellman, Steven A. Rosenberg. Lippincott-Raven Publishers, Philadelphia. (1997)]. Likewise, proangiogenic therapy is directed not only to restoring required angiogenic factors, but to reestablishing the proper balance between them (Dor, et al, Ann N Y Acad Sci 2003; 995:208-16)(for an extensive review of pro- and antiangiogenic therapies see Zhang et al Acta Bioch and Biophys Cinica, 2003: 35:873-880, and Mariani et al. Med Gen Med 2003, 5:22; and Folkman, Semin. Onc 2002, 29:15-18).
Antiangiogenic Therapy:
Anti-angiogenic therapy is a robust clinical approach, as it can delay the progression of tumor growth (e.g., retinopathies, benign and malignant angiogenic tumors).
In general, every disease caused by uncontrolled growth of capillary blood vessels such as diabetic retinopathy, psoriasis, arthritis, hemangiomas, tumor growth and metastasis is a target for anti-angiogenic therapy.
For example, the progressive growth of solid tumors beyond clinically occult sizes (e.g., a few mm3) requires the continuous formation of new blood vessels, a process known as tumor angiogenesis. Tumor growth and metastasis are angiogenesis-dependent. A tumor must continuously stimulate the growth of new capillary blood vessels to deliver nutrients and oxygen for the tumor itself to grow. Therefore, either prevention of tumor angiogenesis or selective destruction of tumor's existing blood vessels (vascular targeting therapy) underlies anti-angiogenic tumor therapy.
Recently, a plethora of anti-angiogenic agents has been developed for the treatment of malignant diseases, some of which are already under clinical trials (for review see Herbst et al. (2002) Semin. Oncol. 29:66-77, and Mariani et al, Med Gen Med 2003; 5:22).
The most studied target for tumor anti-angiogenic treatment is the dominant process regulating angiogenesis in human i.e., the interaction of vascular endothelial growth factor (VEGF) with its receptor (VEGFR). Agents which regulate VEGFR pro-angiogenic action include (i) antibodies directed at the VEGF protein itself or to the receptor (e.g., rhuMAb VEGF, Avastin); (ii) small molecule compounds directed to the VEGFR tyrosine kinase (e.g., ZD6474 and SU5416); (iii) VEGFR targeted ribozymes.
Other novel angiogenesis inhibitors include 2-Methoxyestradiol (2-ME2) a natural metabolite of estradiol that possesses unique anti-tumor and anti-angiogenic properties and angiostatin and endostatin—proteolytic cleavage fragments of plasminogen and collagen XVIII, respectively.
Though promising in pre-clinical models, to date systemic administration of all anti-angiogenic agents tested in clinical trials, have shown limited rate of success and considerable toxicities including thrombocytopenia, leukopenia and hemoptysis. These results suggest that there may be limits to the use of current tumor anti-angiogenic agents as therapy for advanced malignancies. O'Reilly et al. have shown that the latency between the initiation of anti-angiogenic therapy and antitumor effect may result in initial tumor progression before response to therapy [O'Reilly S et al. (1998) Proc Am Soc Clin Oncol 17:217a]. Furthermore, recent studies suggest that the regulation of angiogenesis may differ among capillary beds, suggesting that anti-angiogenic therapy may need to be optimized on an organ/tissue-specific basis [Arap et al. (1998) Science 279:377-380].
Interestingly, poor results have also been obtained when anti-angiogenic therapy (e.g., heparin, heparin-peptide treatment) directed at smooth muscle cell proliferation has been practiced on myocardial ischemia in patients with coronary artery disease [Liu et al., Circulation, 79: 1374-1387 (1989); Goldman et al., Atherosclerosis, 65: 215-225 (1987); Wolinsky et al., JACC, 15 (2): 475481 (1990)]. Various limitations associated with the use of such agents for the treatment of cardiovascular diseases included: (i) systemic toxicity creating intolerable level of risk for patients with cardiovascular diseases; (ii) interference with vascular wound healing following surgery; (iii) possible damage to surrounding endothelium and/or other medial smooth muscle cells.
Thus, these and other inherent obstacles associated with systemic administration of anti-angiogenic factors (i.e., manufacturing limitations based on in-vitro instability and high doses required; and peak kinetics of bolus administration attributing to sub-optimal effects) limit the effective use of angiogenic factors in treating neo-vascularization associated diseases.
Anti-Angiogenic Gene Therapy for Cancer
Tumor cell proliferation in primary tumors as well as in metastases is offset by an increased rate of apoptosis due to a restricted supply of nutrients. Dormant primary or metastatic tumors begin to develop metastases whenever an “angiogenic switch” occurs and nutrient supply is adequate for the size of the tumor.
An angiogenic switch may occur via several mechanisms:
1. Up-regulation of pro-angiogenic genes such as VEGF and bFGF by oncogenes, or down-regulation of angio-suppressors such as thrombospondin.
2. Activation of hypoxic inducible factor-1 (HIF-1) by tumor-related hypoxic conditions.
3. Pro-angiogenic protein secretion by tumor bed fibroblasts, which are induced by tumor cells.
4. Bone marrow endothelial progenitors trafficking to the tumor.
TABLE 1Endogenous regulators of tumor angiogenesis.*Dose-dependent**Weak Angiogenic Activator
The relative balance between activators and inhibitors of angiogenesis (see Table 1 hereinabove) is important for maintaining tumors in a quiescent state. Reducing inhibitors or increasing activator levels alters the balance and leads to tumor angiogenesis and tumor growth.
Oxygen diffusion to neoplastic tissue is inadequate when tumor tissue thickness exceeds 150-200 μm from the nearest vessel. So, by definition, all tumors that exceed these dimensions are already angiogenically switched-on. The tumor cell proliferation rate is independent of the vascular supply. However, as soon as the angiogenic switch occurs, the rate of apoptosis decreases by 3-4 fold (24). Furthermore, nutrient supply and catabolite release are not the only contribution of angiogenic vessels to the decline in tumor apoptosis. Microvasculature endothelial cells also secrete anti-apoptotic factors, mitogens and survival factors such as b-FGF, HB-EGF, IL-6, G-CSF, IGF-1 and PDGF that further suppress tumor cell apoptosis.
Tumor cells are genetically unstable due to high mutation rates, which provide them with an advantage over native cells. For example, mutations in the p53 gene suppress the rate of apoptosis. Moreover, oncogene alteration of pro-angiogenic or angiogenic suppressor control (such as the ras oncogene) may induce an angiogenic switch. However, a high mutation rate is not the only mechanism for cancer's genetic instability. There is evidence of “apoptotic bodies” phagocytosed by tumor cells, resulting in aneuploidy and a further increase in genetic instability. All in all, cancer relies on angiogenesis. Due to genetic instability, cancer may orchestrate a pro-angiogenic cytokine balance, which suppresses its apoptotic rate and enables metastatic seeding.
The human vasculature system contains more than one trillion endothelial cells. The lifetime of normal quiescent endothelial cells exceeds 1000 days. Although angiogenic endothelial cells involved in tumor progression proliferate rapidly, they differ from tumor cells by their genomic stability, and thus also in minimal drug resistance and low likelihood of the development of mutant clones. Moreover, since the rate-limiting factor for tumor progression is angiogenesis, treatment directed against angiogenic endothelial cells could yield highly effective treatment modalities. Indeed, several anti-angiogenic substances could serve as potential candidates for systemic therapy. However, since these agents are proteins and their administration therefore depends on frequent intravenous administration, their use poses serious manufacturing and maintenance difficulties. Delivery of anti-angiogenic genes offers a potential solution for continuous protein secretion.
With the identification of new genes that regulate the angiogenic process, somatic gene therapy has been attempted to overcome these limitations. Although, great efforts have been directed towards developing methods for gene therapy of cancer, cardiovascular and peripheral vascular diseases, there is still major obstacles to effective and specific gene delivery [for review see, Feldman AL. (2000) Cancer 89(6):1181-94] In general, the main limiting factor of gene therapy with a gene of interest, using a recombinant viral vector as a shuttle is the ability to specifically direct the gene of interest to the target tissue.
Attempts to overcome these limitations included the use of tissue-specific promoters conjugated to cytotoxic genes. For example, endothelial cell targeting of a cytotoxic gene, expressed under the control endothelial-specific promoters has been described by Jagger et al who used the KDR or E-selectin promoter to express TNFα specifically in endothelial cells [Jaggar RT. Et al. Hum Gene Ther (1997) 8(18):2239-47]. Ozaki et al used the von-Willebrand factor (vWF) promoter to deliver herpes simplex virus thymidine kinase (HSV-tk) to HUVEC [Hum Gene Ther (1996) 7(13):1483-90]. However, these promoters showed only weak activity and did not allow for high levels of expression.
Several endothelial cell specific promoters have been described in the prior art. For example, Aird et al., [Proc. Natl. Acad. Sci. (1995) 92:7567-571] isolated 5′ and 3′ regulatory sequences of human von Willebrand factor gene that may confer tissue specific expression in-vivo. However, these sequences could mediate only a heterogeneous pattern of reporter transgene expression. Bacterial LacZ reporter gene placed under the regulation of von Willebrand regulatory elements in transgenic mice revealed transgene expression in a subpopulation of endothelial cells in the yolk sac and adult brain. However, no expression was detected in the vascular beds of the spleen, lung, liver, kidney, heart, testes and aorta as well as in the thrombomodulin locus.
Korhonen J et al [Blood (1995) 96:1828-35] isolated the human and mouse TIE gene promoter which contributed to a homogeneous expression of a transgene throughout the vascular system of mouse embryos. However, expression in adult was limited to the vessels of the lung and kidney and no expression was detected in the heart, brain, liver. Similar results were obtained by Schlaeger M et al. who isolated a 1.2 kb 5′ flanking region of the TIE-2 promoter, and showed transgene expression limited to endothelial cells of embryonic mice [Schlaeger T M et al. (1995) Development 121:1089-1098].
Thus, none of these sequences work uniformly in all endothelial cells of all developmental stages or in the adult animal. Furthermore, some of these sequences were not restricted to the endothelium.
An alternate approach presented by Kong and Crystal included a tumor specific expression of anti-angiogenic factors. To date, however, the toxicity of recombinant forms of endogenous anti-angiogenic agents has not been demonstrated although some synthetic anti-angiogenic agents have been associated with toxicity in preclinical models [Kong and Crystal (1998) J. Natl. Cancer Inst. 90:273-76].
Angiostatin has also been used as a possible anti-angiogenic agent (Folkman et al, Cell 1997 Jan. 24;88(2):277-85), however due to the redundancy of factors involved in regulation of angiogenesis in tumors, it is highly unlikely that angiostatin therapy alone would be effective.
To date, promising clinical trials have shown that anti angiogenic treatments like Avastin® or Bay-43906®, can slow the metastatic progression by limiting new growth of blood vessels surrounding the tumors. However, inhibiting the formation of new blood vessels and/or partially destroying them may be insufficient in cancer pathologies where a dramatic anti angiogenic effect that destroys most or all existing angiogenic blood vessels and induce tumor necrosis is required.
The Pre-Proendothelin-1 (PPE-1) Promoter
The endothelins (ET), which were discovered by Masaki et al. in 1988, consist of three genes: ET-1, ET-2 and ET-3. Endothelin-1 (ET-1), a 21 amino acid peptide, was first described as a potent vasoconstrictor and smooth muscle cell mitogen, synthesized by endothelial cells. ET-1 is expressed in the vascular endothelium, although there is some expression in other cells such as smooth muscle cells, the airways and gastrointestinal epithelium, neurons and glomerular mesangial cells. Its expression is induced under various pathophysiological conditions such as hypoxia, cardiovascular diseases, inflammation, asthma, diabetes and cancer. Endothelin-1 triggers production and interacts with angiogenic factors such as VEGF and PDGF and thus plays a role in the angiogenic process.
Hu et al. identified a hypoxia responsive element (HRE) that is located on the antisense strand of the endothelin-1 promoter. This element is a hypoxia-inducible factor-1 binding site that is required for positive regulation of the endothelin-1 promoter (of the human, rat and murine gene) by hypoxia. Hypoxia is a potent signal, inducing the expression of several genes including erythropoietin (Epo), VEGF, and various glycolytic enzymes. The core sequence (8 base pairs) is conserved in all genes that respond to hypoxic conditions and the flanking regions are different from other genes. The ET-1 hypoxia responsive element is located between the GATA-2 and the AP-1 binding sites.
Bu et al. identified a complex regulatory region in the murine PPE-1 promoter (mET-1) that appears to confer endothelial cell specific transcriptional activity and to bind proteins or protein complexes that are restricted to the endothelial cell. This region, designated endothelial specific positive transcription element, is composed of at least three functional elements, positioned between the −364 bp and −320 bp of the murine PPE-1 promoter. All three elements are required for full activity. When one or three copies are constructed into a minimal mET-1 promoter, reporter gene expression in endothelial cells in vitro increased 2-10 times, compared to a minimal promoter with no element.
U.S. Pat. No. 5,747,340 teaches use of the murine PPE-1 promoter and portions thereof. However, this patent neither implies nor demonstrates that an endothelial-specific enhancer can be employed to increase the level of expression achieved with the PPE promoter while preserving endothelial specificity. Further, this patent does not teach that the PPE-1 promoter is induced to higher levels of transcription under hypoxic conditions.
Gene-Directed Enzyme Prodrug Therapy (GDEPT):
This strategy is also called “suicide gene therapy”. It involves the conversion of an inert prodrug into an active cytotoxic agent within the cancer cells. The two most widely used genes in GDEPT are herpes simplex virus thymidine kinase (HSV-TK) coupled with ganciclovir (GCV) administration and the E.coli cytosine deaminase (CD) coupled with 5-fluorocytosine (5FC) administration. The HSV-TK/GCV system has undergone extensive preclinical evaluation, as well as clinical trials. To date, the HSV-TK/GCV system has demonstrated non-significant side effects such as fever, systemic toxicity of GCV, myelosuppression and mild-moderate hepatotoxicity.
The HSV-TK/GCV system was first described by Kraiselburd et al. in 1976. Cells transfected with an HSV-TK containing plasmid or transduced with an HSV-TK containing vector, are becoming sensitive for a super family of drugs including aciclovir, ganciclovir (GCV), valciclovir and famciclovir. The guanosine analog GCV is the most active drug in the setup of gene therapy. HSV-TK positive cells produce a viral TK, which is three orders of magnitude more efficient in phosphorylating GCV into GCV monophosphate (GCV-MP) than the human TK. GCV-MP is subsequently phosphorylated by the native thymidine kinase into GCV diphosphate and finally to GCV triphosphate (GCV-TP).
GCV-TP is a potent DNA polymerase inhibitor leading to termination of DNA synthesis by incorporation into the nascent strand, terminating DNA elongation and eventually causing cell death. Since GCV affects predominantly HSV-TK positive cells, its adverse effects are minimal and rare, and include mainly thrombocytopenia, neutropenia and nephrotoxicity. Moreover, since GCV toxicity is based on DNA synthesis, it affects mainly proliferating cells. The HSV-TK/GCV system has recently been utilized extensively in clinical trials of cancer gene therapy. Nevertheless, results are disappointing, mostly limited in vivo by a low transduction percentage.
Recent studies have characterized the HSV-TK/GCV cell cytotoxicity mechanism. They revealed cell cycle arrest in the late S or G2 phase due to activation of the G2-M DNA damage checkpoint. These events were found to lead to irreversible cell death as well as a bystander effect related to cell death. Profound cell enlargement is a well-known morphological change in cells administered with the HSV-TK/GCV system. These morphological changes are due to specific cytoskeleton rearrangement. Stress actin fibers and a net of thick intermediate filaments appear following cell cycle arrest.
The HSV-TK/GCV system utilizes an amplification potential designated as the “bystander effect”. The bystander effect stands for the phenomenon by which HSV-TK positive cells induce the killing of HSV-TK negative cells.
Bystander effect: The bystander effect was fist described by Moolten et al., who found that a 1:9 mixture of HSV-TK positive and HSV-TK negative cells, respectively, results in complete cell killing following the addition of GCV. Several characteristics of the bystander effect have been described:
1. The bystander effect was found to be highly dependent on cell-cell contact.
2. Its extent was different in different cell types.
3. It is not limited to homogenous cell types, but also to mixtures of different cell types.
4. Higher levels of HSV-TK expression were found to correlate with a higher bystander effect.
Culver et al. were the first to demonstrate a bystander effect in an in vivo model. They demonstrated tumor regression when implanted with HSV-TK positive tumor cells in different ratios. Unlike in vitro models, cell-cell contact was not found to be essential for the bystander effect in vivo. Kianmanesh et al. demonstrated a distant bystander effect by implanting tumor cells in different liver lobes, where only some were HSV-TK positive. Both HSV-TK positive and negative foci regressed. A bystander effect was also demonstrated in vivo between cells from different origins. All in all, HSV-TK and its bystander effect facilitate an effective means for tumor suppression when implemented in gene delivery systems. However, to date, clinical studies have demonstrated only limited results.
There is thus a widely recognized need for, and it would be highly advantageous to have highly specific, reliable angiogenic-specific promoters and nucleic acid constructs providing a novel approach for efficiently regulating angiogenesis in specific tissue regions of a subject while being devoid of the toxic side effects and limited success characterizing prior art anti-angiogenesis approaches.